Deep learning-assisted self-cleaning cellulose colorimetric sensor array for monitoring black tea withering dynamics

Area of Science:

• Materials Science
• Analytical Chemistry
• Food Science

Background:

• The withering process significantly impacts black tea's aroma and quality.
• Accurate monitoring of tea withering stages is crucial for consistent product development.
• Existing methods for assessing tea quality can be labor-intensive and subjective.

Purpose of the Study:

• To develop an eco-friendly, cellulose film-based colorimetric sensor array (CSA) for detecting volatile organic compounds (VOCs).
• To assess the withering stages of black tea using the developed CSA and deep learning.
• To introduce a sustainable fabrication strategy for smart sensor applications.

Main Methods:

• Fabrication of a self-cleaning TiO2-cellulose film via TiO2 attachment.
• Creation of functionalized cellulose film with hydrophobic non-sensing areas using octadecyltrichlorosilane (OTS).
• Preparation of the OTS/TiO2-CSA by drop-coating dyes onto hydrophilic sensing areas, enhancing humidity resistance.
• Utilizing a Long Short-Term Memory deep learning model for data analysis and stage assessment.

Main Results:

• The fabricated OTS/TiO2-CSA demonstrated improved humidity resistance.
• The sensor array, combined with a deep learning model, achieved 90% accuracy in identifying tea withering stages.
• The sensor dyes exhibited significant degradation (>70%) under limited UV exposure, indicating potential for controlled use or disposal.

Conclusions:

• A smart, eco-friendly OTS/TiO2-CSA was successfully fabricated using a novel strategy.
• The developed sensor array shows significant potential as a sustainable tool for real-time monitoring of tea withering.
• This approach offers a promising advancement in quality control for the black tea industry.